This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Purpose of DAC Renewal The aim of our ongoing computational fluid dynamics (CFD) studies are i) to understand and quantify the suboptimal physiological state of the failing Fontan patient group as an extension of our previous study that identified the caval waveform driven energetics of the functionally healthy patients [1, 2], ii) to characterize the hemodynamic loading at the intermediate embryonic development time points as an extension of the baseline data that was recently documented by our group [3, 4] iii) to design interactive surgical planning methods to enable cyber generated intelligence for aiding surgical decision making process in time critical, patient specific coronary artery bypass (CABG) operations [5]. Resources provided during our initial grant period supported two in press and two submitted journal publications. In addition our results are featured and PSC resources are fully acknowledged in 6 conferences and invited presentations. Most importantly the simulations enabled a competitive NSF-CDI proposal prepared by the joint effort of CMU engineering faculty (Pekkan-BME, Kara-ME, Zhang-ME). Introduction: Summary of studies undertaken during previous DAC grant period With the advance of accurate anatomical reconstruction modalities (via magnetic resonance imaging, angiograms or computational tomography) and robust numerical schemes, CFD simulations has been a viable tool for surgical planning and device design in cardiovascular medicine [6-14]. CFD allow not only quantifying the local hemodynamics inside patient-specific anatomies but also to evaluate and optimize the performance of candidate surgical design templates [15-18]. A typical systematic pre-surgical planning scenario incorporating large number of geometrical variations and non-intuitive design alternatives require high performance computing power as the raw data of such an optimization study alone allocates ~2.4Terabytes of temporary and permanent data storage systems [19, 20]. Energy dissipation in failing Fontans: Last stage of the palliative surgical reconstruction (i.e. Fontan procedure) for the infants with functional single-ventricle is total cavopulmonary connection (TCPC), where the superior vena cavae (SVC) and inferior vena cavae (IVC) are routed directly into the pulmonary arteries. Limited pumping energy available due to the absence of right-ventricle and altered venous characteristics require optimized hemodynamics inside the TCPC pathway, which can be achieved by minimizing the power losses. Earlier studies on Fontan hemodynamics exclusively focused on patients who are doing well despite the growing number of adult late survivors with declining hemodynamic states [21]. Hence, the energy state in the latter group has not been documented. During our previous Development Allocation (DAC) grant provided by Pittsburgh Supercomputing Center (PSC) we discovered that caval flow waveforms have significant impact on pulsatile energetics (i.e. 2-34% power loss variation) inside the TCPC pathway for the fixed geometry [1]. Along this line, the caval waveform topologies of failing Fontan patients are considerably different than the baseline healthy flow waveforms. Therefore, quantification of the energy efficiency of the failing waveform topology may provide an additional hemodynamic parameter that can correlate with cardiac malfunction and postoperative complications [22-24]. In addition, energetic state of failing Fontans may also shed light to the consequences of the phrenic nerve injury (a surgical remnant in Fontan surgery) since despite the surgical plication these defects pose similar perturbed hemodynamics such as those seen in failing Fontan circulation [25]. Embryonic aortic arch development: Heart and vascular abnormalities comprise a vital category of human birth defects, occurring in approximately 1% of live born infants [26]. Genetic and epigenetic (environmental and biomechanical) factors regulate the cardiovascular (CV) morphogenesis and adaptation which involves dynamic three-dimensional (3D) changes in CV structure and function [27, 28]. Flow-driven hemodynamics plays a significant role in this dynamic process which is clinically described as a flow-dependency principle [29-31]. Altered intrauterine hemodynamics during critical windows of cardiac morphogenesis and remodeling leads to congenital arterial defects, such as interruption of aortic arch, double aortic arches, and coarctation of aorta [32, 33]. While the early origins and progression of these defects are well established, little is known regarding the relationships between individual arch geometry and hemodynamics, the distribution of cardiac output through each of the arches, and the influence of altered hemodynamic forces on dynamic arch remodeling. During our previous DAC grant awarded by PSC, we quantified the changes in the three-dimensional (3D) geometry of aortic arch, blood flow, and shear stress patterns between two Hamburger-Hamilton (HH) developmental stages i.e. HH18 - HH24 (where both embryonic mass and cardiac output quadruples) of normal arch morphogenesis. This initial study documents the flow distribution and hemodynamic loading inside the normal developing aortic arch for the first time in the literature. Hence, it provides an important baseline data for our ongoing studies to investigate congenitally defected models created via microsurgery techniques (i.e. left-atrium ligation to simulate hypoplastic left heart syndrome) in our laboratory (Fig.1). Improved patient-specific CABG configurations: Statistics by the American Heart Association identify coronary artery disease (CAD) as the primary cause of morbidity and mortality in the western world [34]. As a common invasive treatment to CAD, bypass conduits provide an alternative route around critically blocked arteries. Current surgical anastomosis techniques and the design of synthetic CABG frequently lead to post-surgical complications such as intimal thickening, restenosis and eventual long term graft failure [35]. Pathological hemodynamic states are known as the precursors of intimal hyperplasia or platelet deposition and result in graft occlusion. From fluid mechanics perspective, abnormalities in coronary flow, which include recirculation zones, low/oscillating shear stresses, vortices, and areas of stagnation, have to be reduced to achieve a failure-free CABG design. Hence, the current CABG design paradigm aims improved hemodynamics, to achieve reduced hyperplasia at the distal anastomosis region by modulating the anastomosis angle and the vessel curvature in simplified 2D tubular conduits. Although in-plane (2D) optimization is appropriate to identify these primary design features, an accurate assessment on the CABG hemodynamics requires patient-specific 3D anatomical information for reliable feedback. During our previous DAC grant period we embark upon developing a CFD coupled multi-dimensional shape optimization method to aid patient specific CABG design. Supported by the high performance computing (HPC) resources, this framework will allow us to analyze hemodynamics efficiency of end-to-side anastomosis to left anterior descending artery (LAD) and sequential grafting strategies, i.e. raising multiple (parallel) braches from a parent CABG. Example Methodology Energy dissipation in failing Fontans: Numerical simulations are performed using the idealized 1DO standard TCPC geometry in order to illustrate the isolated effect of flow waveform topology on power loss [1]. CFD model incorporates the experimentally validated unsteady 2nd order accurate solver [10, 14] in FLUENT version 6.3.26 (ANSYS Inc., Canonsburg, PA) to simulate incompressible and Newtonian blood flow with constant hemodynamic properties (? = 1060 kg/m3, ? = 3.71 10-3 Pa.s). The power dissipation inside the circulation is calculated using the control volume approach neglecting the heat losses and rate of work done by the walls. Failing Fontan caval waveforms are acquired in-house at University of Pittsburgh, Childrens Hospital through an approved IRB using multichannel simultaneous ECG, respiration and real-time ultrasound measurements. Reconstructed waveforms for a patient with severe suboptimal hemodynamics (NYHA functional class III, Age:16, BSA: 1.2, HR:56, respiration rate: 32/min) for 3 respiratory cycles is given in Figure 2. To compare the relative energy efficiency corresponding functional Fontan waveforms are reproduced from the patient-specific real-time PC-MRI caval waveform data set available in literature [36]. Due to the severe retrograde flow observed in both data sets the outflow mass split boundary condition that is introduced in our previous work (for unidirectional caval flow) [10] is found to be incapable of preserving the mass conservation in the presence of backflow from the outlets and improved here. Instead, a user-defined pressure boundary condition subroutine, incorporating a numerical scheme derived from the Bernoulli equation is assigned to the pulmonary outlets. At each iteration this numerical scheme adjusts the pulmonary outlet pressure incrementally in order to reach the specified 50/50 baseline flow split value between left and right lung. Systematic verification tests and characterization of this new outlet boundary condition that is specifically useful for failed Fontan waveforms with backflow are conducted. Embryonic aortic arch development: Composite 3D models of the chick embryo aortic arches were generated at the Hamburger-Hamilton (HH) developmental stages HH18 and HH24 using fluorescent dye injection, micro-CT, Doppler velocity recordings and pulsatile subject-specific computational fluid dynamics (CFD). India ink and fluorescent dyes were injected into the embryonic ventricle or atrium to visualize left and right aortic arch morphologies and flows. 3D morphology of the developing great vessels was obtained from polymeric casting followed by micro-CT scan (Fig.1). The flow domain inside the aortic arches is discretized using unstructured tetrahedral elements (Gambit, Ansys Inc., PA). A detailed grid sensitivity study was conducted to investigate the effects of both surface and core mesh refinements (i.e. 9 auxiliary CFD simulations). Mesh independent solutions were obtained using approximately 600,000 tetrahedral elements at each stage. Mesh generation was performed using Gambit (ANSYS Inc, Canonsburg, PA) and the aforementioned pulsatile 2nd order CFD solver was utilized. Inlet aortic arch flow and body-to-brain flow-split was obtained from 20MHz pulsed Doppler velocity measurements and literature data. Improved patient-specific CABG configurations: The fully automated optimization framework is built upon coupling the CFD based evaluation of the cost function into the optimization algorithm in an automated fashion (Fig.3). User defined subroutines are used to generate desired CABG shape variations and to ensure robust communication between the shape optimizer and the CFD solver (both in house code and commercial codes). LAD CABG initiates from the ascending aorta and anastamosed to the native coronary artery distal to the stenosed region. The proximal and end anastomosis locations are selected according to the surgical settings of an actual proximal LAD grafting performed in Bovine heart (Fig.4). CABG is simplified as a 2D cylindrical tube for the in-plane optimization. The scaffold of the coronary vessel is created using a 3rd order Bezier curve whose shape is dictated by four design parameters namely the proximal (?p) and distal (?d) anastomosis angles, proximal and distal curvature vectors (sp, sd). Grid independency of the solution is ensured by comparing the solutions at six refinement levels. Steady state CFD solver of Fluent (ANSYS Inc., Canonsburg, PA) is used to solve the governing Navier-Stokes equations for each geometric alteration. A parabolic velocity profile is prescribed at the inlet boundary of the graft to achieve more realistic coronary flow. Once the optimized 2D graft shape is obtained, it is translated to 3D environment by using the in-house anatomical editing tool. Efficiency of the final graft design is evaluated using experimentally validated 2nd order CFD solver of Fluent (ANSYS Inc, PA) incorporating impedance based outlet boundary conditions tuned to simulate coronary blood flow. Preliminary Results Energy dissipation in failing Fontans: For the functional Fontan data power loss calculated for one respiratory cycle returns 9.82 mW, whereas, power loss in failing Fontan patients is 11.3 mW which is 15% higher than the power loss calculated for functional Fontan patient. Having higher power loss values at this critical condition create further reductions in cardiac output [37]. A detailed comparison of the time averaged flow structures indicates that the axial flows are increased and the vorticity of the secondary flows are significantly amplified in the failing Fontan waveform configuration (Fig.5). Hence, higher power loss value is associated with the increase in the rotational strength of the counter-rotating vortices along the PAs. Embryonic aortic arch development: Figure 6 shows the distribution of wall shear stress (WSS) on the left and right laterals of HH18 and HH24 at peak cardiac phase. It is found that WSS was high at the flow divider and narrowing parts and distribution is complex in the vicinity of aortic sac. Time-averaged WSS was calculated for representative cardiac cycles at HH18 and HH24 and found to be 15.5 dynes cm-2 and 28.5 dynes cm-2, respectively. Increased WSS correspond to increased cardiac output at HH24 relative to HH18. Average lateral time-averaged WSS correlated with average mean midpoint diameter of the three aortic arches at HH18 and at HH24. Average right lateral midpoint diameter of the aortic arches at HH18 (108 0.022 ?m) was significantly larger than the average left lateral diameter (101 0.015 ?m, p<0.05). Likewise, the higher peak and mean lateral time-averaged WSS (16 dynes cm-2 and 55 dynes cm-2 respectively, p<0.05) at HH18 were observed in the right sided arches versus the left sided arches in the CFD model. Similar conclusions also apply for development stage HH24. Improved patient-specific CABG configurations: Preliminary results indicate that the 2D optimization scheme attempts to alter the CABG geometry to minimum curvature and minimum length configuration (Fig.6). Hence, an arbitrary highly curved graft topology transforms into a relatively straight configuration by minimizing the magnitude of the end tangent vectors and anastomosis angles. In addition, the optimization results for the given problem settings indicate that the energy dissipation and vorticity inside the graft decreases about 20% from suboptimal to optimal configurations. On the other hand, LAD flow rate calculated through steady state 3D simulations of native coronary circulation agrees well with the previous ultrasound angiography measurements [38]. Concluding Remarks and Future Research Aims: Energy dissipation in failing Fontans: This study is a first attempt towards investigating and quantifying hemodynamics in failing Fontan patients. Higher energy losses in 1DO TCPC found for the analyzed failing Fontan patient compared to the functional Fontan at the same normalized cardiac output originates due to the fluctuating IVC and SVC waveforms. Since the influence of TCPC hydrodynamic power loss on cardiac output is quite significant (sensitivity = -0.88 L/Min/Woods Unit) [37], the energy efficiency of the caval waveforms can be correlated with cardiac function and further be used in planning the post-operative management of patients with declining hemodynamics. Future efforts will expand the limited real-time data through additional clinical studies and utilize HPC to perform CFD simulations within realistic patient-specific anatomical templates for improved understanding of Fontan energetics throughout the disease timeline. Due to the patient-to-patient variations in cardiopulmonary interactions (i.e. variations in caval flow waveforms) a large patient population (>10 patient from each group) is required for making a quantitative assessment on the energetic state of adult failing Fontan patients. Hence, we estimate the total row data of this CFD modeling effort about ~2 Terabytes of storage. Embryonic aortic arch development: Expanding on the previous intra-cardiac hemodynamic investigations, the present study is a first attempt to reconstruct the normal (healthy) micro 3D aortic arch morphology and flow distribution during a critical period of development based on anatomical casts in chick embryo. Between stages HH18 and HH24 extensive 3D anatomical changes in size, curvature and orientation of the developing arches and cranial vessels indicate that the intermediate stage, HH21 may be the key stage in aortic arch development. Our ongoing microsurgeries and cast formations indicate that HH21 pose rapid, instable morphological changes and multiple arch modalities. Therefore, with the renewal of DAC from PSC we aim to characterize the intriguing hemodynamics of complex HH21 stage (~50 embryonic cast reconstructions;~0.5-1 Terabytes row data storage) where both the genetic and epigenetic factors could result in abrupt morphological changes including the initiation of congenital arch malformations. Improved patient-specific CABG configurations: This study illustrates a fully automated efficient framework for coupling optimal shape design to 2D non-Newtonian blood flow simulations in idealized cardiovascular geometries. As a major outcome, the clinical interpretation of less power loss inside the CABG may correlate with increased coronary perfusion based on the decreased vascular resistance of the graft geometry. Therefore, for the first time this study identifies the significance of the bulk shape in the CABG design that has long been overlooked. Future studies under pulsatile coronary flow settings will investigate the importance of vorticity and oscillating shear stress. These flow parameters correlate strongly with the vasoregulation and disease states [39]. Under transient flow conditions, optimization paradigm incorporating a large design space (i.e. 20-100 templates) to account for the anatomically correct optimal CABG design requires high computational power (~1 Terabyte per single CABG design). Our ongoing efforts will parallelize the proposed 2D optimization paradigm to incorporate HPC resources and investigate the hemodynamic efficiency of 3D sequential bypass graft configurations (Fig.7). In conclusion, our previous DAC from PSC enabled us to excel our skills in managing HPC resources and provided the necessary tools to advance our cutting-edge cardiovascular research. Although initially we are awarded with allocations on the Rachel system, we conducted majority of our computations via our friendly-user-account on Pople (i.e. without using the computation time allocated with the grant) which was generously provided to us by PSC. To continue our ongoing studies we are, once again, looking for high computing power which is beyond the limits of that is currently eligible through our university. References: [1] Dur, O., DeGroff G, C., and Pekkan, K., 2009, "Optimization of inflow waveform topology for minimized total cavopulmonary power loss," Journal of Biomechanical Engineering. [2] Dur, O., Sundareswaran, K., DeGroff, C., Yoganathan, A., and Pekkan, K., 2008, "Optimization of Total Cavopulmonary Power Loss via Caval Flow Waveforms based on Patient Specific Real Time Echo and MRI Data," Biomedical Engineering Symposium St. Louis, MO. [3] Pekkan, K., Dasi, L. P., Nourparvar, P., Yerneni, S., Tobita, K., Fogel, M. A., Keller, B., and Yoganathan, A., 2008, "In vitro hemodynamic investigation of the embryonic aortic arch at late gestation," J Biomech, 41(8), pp. 1697-1706. [4] Wang, Y., Dur O., Patrick, M., Tinney J., Tobita K., Keller, B., and Pekkan, K., 2009, "Aortic arch morphogenesis and flow modeling in the chick embryo," Annals of Biomedical Engineering. [5] Dur, O., Coskun, S., Coskun, K., Kara, L., and Pekkan, K., 2009, "Improved Patient-Specific Coronary Artery Graft Configurations using CFD Coupled Shape Optimizer," ASME, Summer Bioengineering Conference Lake Tahoe, CA. [6] Bove, E. L., de Leval, M. R., Migliavacca, F., Guadagni, G., and Dubini, G., 2003, "Computational fluid dynamics in the evaluation of hemodynamic performance of cavopulmonary connections after the Norwood procedure for hypoplastic left heart syndrome," J Thorac Cardiovasc Surg, 126(4), pp. 1040-1047. [7] Burgreen, G. W., Antaki, J. F., Wu, Z. J., and Holmes, A. J., 2001, "Computational fluid dynamics as a development tool for rotary blood pumps," Artif Organs, 25(5), pp. 336-340. [8] DeGroff, C., Birnbaum, B., Shandas, R., Orlando, W., and Hertzberg, J., 2005, "Computational simulations of the total cavo-pulmonary connection: insights in optimizing numerical solutions," Med Eng Phys, 27(2), pp. 135-146. [9] Pekkan, K., de Zelicourt, D., Ge, L., Sotiropoulos, F., Frakes, D., Fogel, M. A., and Yoganathan, A. P., 2005, "Physics-driven CFD modeling of complex anatomical cardiovascular flows-a TCPC case study," Ann Biomed Eng, 33(3), pp. 284-300. [10] Pekkan, K., Dur, O., Sundareswaran, K., Kanter, K., Fogel, M., Yoganathan, A., and Undar, A., 2008, "Neonatal aortic arch hemodynamics and perfusion during cardiopulmonary bypass," J Biomech Eng, 130(6), p. 061012. [11] Pekkan, K., Kitajima, H. D., de Zelicourt, D., Forbess, J. M., Parks, W. J., Fogel, M. A., Sharma, S., Kanter, K. R., Frakes, D., and Yoganathan, A. P., 2005, "Total cavopulmonary connection flow with functional left pulmonary artery steno angioplasty and fenestration in vitro," Circulation, 112(21), pp. 3264-3271. [12] Scotti, C. M., Jimenez, J., Muluk, S. C., and Finol, E. A., 2008, "Wall stress and flow dynamics in abdominal aortic aneurysms: finite element analysis vs. fluid-structure interaction," Comput Methods Biomech Biomed Engin, 11(3), pp. 301-322. [13] Torii, R., Wood, N. B., Hughes, A. D., Thom, S. A., Aguado-Sierra, J., Davies, J. E., Francis, D. P., Parker, K. H., and Xu, X. Y., 2007, "A computational study on the influence of catheter-delivered intravascular probes on blood flow in a coronary artery model," J Biomech, 40(11), pp. 2501-2509. [14] Wang, C., Pekkan, K., de Zelicourt, D., Horner, M., Parihar, A., Kulkarni, A., and Yoganathan, A. P., 2007, "Progress in the CFD modeling of flow instabilities in anatomical total cavopulmonary connections," Ann Biomed Eng, 35(11), pp. 1840-1856. [15] Abraham, F., Behr, M., and Heinkenschloss, M., 2005, "Shape optimization in steady blood flow: a numerical study of non-Newtonian effects," Comput Methods Biomech Biomed Engin, 8(2), pp. 127-137. [16] Agoshkov, V., Quarteroni, A., and Rozza, G., 2006, "A Mathematical Approach in the Design of Arterial Bypass Using Unsteady Stokes Equations," Journal of Scientific Computing, 28(2). [17] Marsden, A., Feinstein, J., and Taylor, C., 2008, "A computational framework for derivative-free optimization of cardiovascular geometries," Comput. Methods Appl. Mech. Engrg, 197, pp. 18901905. [18] Weiguang Yang, Jeffrey Feinstein, V. Mohan Reddy, and Marsden, A., 2008, "Optimization of an idealized Y-Shaped Extracardiac Fontan Baffle," 61st Annual Meeting of the APS Division of Fluid DynamicsSan Antonio, Texas. [19] Payli, R., Pekkan, K., Zelicourt, D., Frakes, D., Sotiropoulos, F., and Yoganathan, A., 2007, "High Performance Clinical Computing on theTeraGrid: Patient-Specific Hemodynamic Analysis and Surgical Planning," TeraGrid 2007 ConferenceMadison, WI. [20] Pekkan, K., Whited, B., Kanter, K., Sharma, S., de Zelicourt, D., Sundareswaran, K., Frakes, D., Rossignac, J., and Yoganathan, A. P., 2008, "Patient-specific surgical planning and hemodynamic computational fluid dynamics optimization through free-form haptic anatomy editing tool (SURGEM)," Med Biol Eng Comput, 46(11), pp. 1139-1152. [21] DeGroff, C. G., 2008, "Modeling the Fontan circulation: where we are and where we need to go," Pediatr Cardiol, 29(1), pp. 3-12. [22] Anderson, P. A., Sleeper, L. A., Mahony, L., Colan, S. D., Atz, A. M., Breitbart, R. E., Gersony, W. M., Gallagher, D., Geva, T., Margossian, R., McCrindle, B. W., Paridon, S., Schwartz, M., Stylianou, M., Williams, R. V., and Clark, B. J., 3rd, 2008, "Contemporary outcomes after the Fontan procedure: a Pediatric Heart Network multicenter study," J Am Coll Cardiol, 52(2), pp. 85-98. [23] Ghanayem, N. S., Berger, S., and Tweddell, J. S., 2007, "Medical management of the failing Fontan," Pediatr Cardiol, 28(6), pp. 465-471. [24] Marino, B. S., 2002, "Outcomes after the Fontan procedure," Curr Opin Pediatr, 14(5), pp. 620-626. [25] Hsia, T. Y., Khambadkone, S., Bradley, S. M., and de Leval, M. R., 2007, "Subdiaphragmatic venous hemodynamics in patients with biventricular and Fontan circulation after diaphragm plication," J Thorac Cardiovasc Surg, 134(6), pp. 1397-1405;discussion 1405. [26] Sadler, T., and Langman, J., 2006, Langman's Medical Embryology, Lippincott Williams &Wilkins. [27] Hove, J. R., Koster, R. W., Forouhar, A. S., Acevedo-Bolton, G., Fraser, S. E., and Gharib, M., 2003, "Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis," Nature, 421(6919), pp. 172-177. [28] Poelmann, R. E., Gittenberger-de Groot, A. C., and Hierck, B. P., 2008, "The development of the heart and microcirculation: role of shear stress," Med Biol Eng Comput, 46(5), pp. 479-484. [29] Gardiner, H., Brodszki, J., Eriksson, A., and Marsal, K., 2002, "Volume blood flow estimation in the normal and growth-restricted fetus," Ultrasound Med Biol, 28(9), pp. 1107-1113. [30] Rudolph, A. M., and Heymann, M. A., 1970, "Circulatory changes during growth in the fetal lamb," Circ Res, 26(3), pp. 289-299. [31] Ursell, P. C., Byrne, J. M., Fears, T. R., Strobino, B. A., and Gersony, W. M., 1991, "Growth of the great vessels in the normal human fetus and in the fetus with cardiac defects," Circulation, 84(5), pp. 2028-2033. [32] Hogers, B., DeRuiter, M. C., Gittenberger-de Groot, A. C., and Poelmann, R. E., 1997, "Unilateral vitelline vein ligation alters intracardiac blood flow patterns and morphogenesis in the chick embryo," Circ Res, 80(4), pp. 473-481. [33] Hogers, B., DeRuiter, M. C., Gittenberger-de Groot, A. C., and Poelmann, R. E., 1999, "Extraembryonic venous obstructions lead to cardiovascular malformations and can be embryolethal," Cardiovasc Res, 41(1), pp. 87-99. [34] Rosamond, W., Flegal, K., Furie, K., Go, A., Greenlund, K., Haase, N., Hailpern, S. M., Ho, M., Howard, V., Kissela, B., Kittner, S., Lloyd-Jones, D., McDermott, M., Meigs, J., Moy, C., Nichol, G., O'Donnell, C., Roger, V., Sorlie, P., Steinberger, J., Thom, T., Wilson, M., and Hong, Y., 2008, "Heart disease and stroke statistics--2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee," Circulation, 117(4), pp. e25-146. [35] Quarteroni, A., and Rozza, G., 2003, "Optimal Control and Shape Optimization of Aorto-Coronaric Bypass Anastomoses," Mathematical Models and Methods in Applied Sciences, 13(12), pp. 1801-1823. [36] Hjortdal, V. E., Emmertsen, K., Stenbog, E., Frund, T., Schmidt, M. R., Kromann, O., Sorensen, K., and Pedersen, E. M., 2003, "Effects of exercise and respiration on blood flow in total cavopulmonary connection: a real-time magnetic resonance flow study," Circulation, 108(10), pp. 1227-1231. [37] Sundareswaran, K. S., Pekkan, K., Dasi, L. P., Whitehead, K., Sharma, S., Kanter, K. R., Fogel, M. A., and Yoganathan, A. P., 2008, "The total cavopulmonary connection resistance: a significant impact on single ventricle hemodynamics at rest and exercise," Am J Physiol Heart Circ Physiol, 295(6), pp. H2427-2435. [38] Manning, W. J., Li, W., and Edelman, R. R., 1993, "A preliminary report comparing magnetic resonance coronary angiography with conventional angiography," N Engl J Med, 328(12), pp. 828-832. [39] Loth, F., Jones, S. A., Zarins, C. K., Giddens, D. P., Nassar, R. F., Glagov, S., and Bassiouny, H. S., 2002, "Relative contribution of wall shear stress and injury in experimental intimal thickening at PTFE end-to-side arterial anastomoses," J Biomech Eng, 124(1), pp. 44-51.